VideoRAG
[KDD'2026] "VideoRAG: Chat with Your Videos"
What is VideoRAG?
[KDD'2026] "VideoRAG: Chat with Your Videos"
Overview of VideoRAG
What it is
VideoRAG is a research framework and the foundation of Vimo, a desktop application that enables users to chat with their videos using AI. Vimo can process videos of any length, from short clips to hundreds of hours of content, and answer questions with high accuracy. The project is open source and includes a benchmark dataset called LongerVideos with over 134 hours of video content.
Key features
- Drag & Drop Upload: Simply drag video files into Vimo
- Smart Conversations: Ask questions in natural language
- Multi-Format Support: Works with MP4, MKV, AVI, and more
- Extreme Long Videos: Process videos up to hundreds of hours
- Multi-Video Analysis: Compare and analyze multiple videos simultaneously
- Advanced Retrieval: Find specific moments and scenes with precision
- VideoRAG Framework: Access to cutting-edge retrieval-augmented generation
Use cases
- Video enthusiasts and professionals can upload any video and start asking questions immediately, with no length limits and deep understanding that combines visual content, audio, and context.
- Researchers and developers can use the state-of-the-art VideoRAG algorithm, which features graph-driven knowledge indexing and is evaluated on 134+ hours across lectures, documentaries, and entertainment.
- The application is cross-platform (macOS, Windows, Linux) and supports multi-video analysis for comparing content.
Getting started
See the repository README for detailed setup instructions.
License and pricing
The license is not stated (NOASSERTION). The project is open source.
Based on the project's public README and repository metadata.